Abstract

Due to the volume change accompanying the fcc to bcc or bct crystal structures in steels, it is a common practice to determine phase transformation temperatures using dilatometry. The martensite start temperature (Ms) is often of particular interest. Experimentally, it is found that the start of the martensite transformation is not indicated by a sharp change in the slope of the dilatation curve as is predicted by the Koistinen–Marburger equation. Rather, there is a gradual change in the slope such that the martensite start temperature is ill-defined. The current work shows that this gradual change in slope can be related to chemical inhomogeneity in the steel caused by interdendritic microsegregation. It is shown that combining the Koistinen–Marburger equation with measured concentration profiles allows experimental dilatation curves to be well predicted.